This proposal is for a continuation of our current 5-year, NIAAA study, which prospectively examines the effect of retirement on the drinking behavior of workers. The purpose of the research is to examine how different levels of social isolation (i.e., depth and breadth of social support or lack thereof), social control (loss of [or relief from] work based systems or rules governing drinking;permissive drinking norms), stress (e.g., financial insecurity, role underload, marital strain), and social marginalization (e.g., lowered self efficacy, loss of self esteem) experienced during the retirement process impact drinking behavior. The study is unique because, in contrast to other studies, it follows retirement-eligible workers (i.e., 6 months prior to their eligibility for retirement) into their retirement years. In the current study, we have identified three different retirement trajectories (e.g., postponing retirement, partial retirement, and full retirement) and begun to analyze their relationship to social isolation, social control, stress, and social marginalization and workers'drinking behavior. During the first 5 years, we proposed to follow retirement-eligible workers from 6 months before they were eligible to retire until 18 months after retirement. At the time of the original submission, we also stated that we planned to submit a continuation proposal, which would permit us to follow our sample of unionized, blue-collar workers into their retirement years. To date, we have collected qualitative data from blue-collar retirees to generate theoretical insight into the impact of the retirement process on drinking behavior as well as 3 waves of survey data from a unique cohort of retirement-eligible, blue-collar workers to test our hypotheses about retirement's impact on drinking behavior and other yet-to-be-identified effects will only become manifested over time. Over the next five years we propose to collect 3 additional waves of survey data from our cohort sample on the basis of an expanded research model;a model incorporating a number of new constructs and hypotheses generated on the basis of our qualitative data. Thus, the continuation will allow us to examine the full effects of retirement as more retirement-eligible workers opt for full or partial retirement and struggle with the implications of their retirement (e.g., isolation, stress, marginalization) and its effects on their drinking behavior,